Continue in 2 seconds

Using Judgment in Business Intelligence

  • June 01 2004, 1:00am EDT

Most business intelligence environments resemble a mishmash (yes, that's the technical term) of well-architected structures and those that were obviously designed to meet an urgent need regardless of the longer-term consequences. This is true for best practice programs, beginning programs and programs that have experienced multiple failures.

Are the well-architected structures right and the others wrong? That depends on your perspective. There is no easy answer (e.g., all independent data marts should never have been created in the first place). However, there are effectiveness and efficiency measures. Effectiveness is measured by the architecture's ability to meet specific needs - and how long it is going to be able to do so with current "support" levels before rearchitecting is necessary. An independent data mart may meet a singular short-term need in a timely and very effective manner; and it may be well-architected when you don't consider the enterprise needs. However, it is efficiency that usually enforces good architecture. Occasionally, application needs cannot be met by the on-the-fly integration required in the absence of an enterprise data warehouse or those applications will be less effective than they could be.

For example, many marketing programs still operate using a data structure developed outside of business intelligence processes and guidelines. As the marketing needs grow, the patchwork approach required to pull together the disparate structures grows too ... and it becomes painful.

Essentially, these departments wish they had an enterprise data warehouse, where all the information they require - both now and in the future - resides, so that they can immediately exploit the data or quickly add it when needed. How long until they bite the bullet and consolidate into an enterprise architecture? It will likely happen, but when? However, if the unarchitected components meet the need accurately and in a timely fashion, it is difficult to argue that the solution is wrong.

The ideals of the corporate information factory are alive and well today. An architecture with multiuse transformations in one place and the availability of a single version of the truth - even though some data may be distributed through a data mart network - is the most efficient approach for supporting enterprise-wide information needs.

Every organization should work from a methodology and architecture approach. However, the approach needs to be tempered with accommodation for urgent information needs. A plan for "getting back" to the desired methodology and architecture should accompany these approach details. Let's call it all "guiding principles."

Guiding principles include the exception conditions for their use. For example, you may have a guiding principle that no data goes to a mart without first traveling through the data warehouse. However, what if an urgent need arises for data in a mart and your data latency factor for navigating through the warehouse structure is too slow to meet the need? You can correct this systemically - as you probably should - but there's no one to step up with the budget for this now. The proponents of the urgent need can argue that is not their burden to bear; and they would be right.

Yet don't give up on your guiding principles too easily. They are worth fighting for. Usually, after adding a few basic processes to the environment, architected solutions are quicker to come by than the unarchitected ones.

This is the essence of a hybrid, best-of-breed approach to business intelligence. It accommodates the urgent information needs of the business while adhering to a flexible, scalable approach that will ultimately provide the most effective balance of both efficiency and effectiveness.

I'm sure most of you can appreciate these mixed messages that are commonplace when advising on areas of judgment. Architecture is a judgment issue - both initially and on an ongoing basis.

The chances of successful business intelligence efforts significantly correlate with having people with the right characteristics for success on the project - or at least enough of them to compensate for those that do not. Also, success goes far beyond technical skills, and good technical skills need to be balanced with communication and direction-setting.

An overriding characteristic for hiring managers for the hybrid, flexible approaches needed today is sound judgment or the ability to arrive at and act on a consensus of opinion, including:

  • The ability to arrive at a rational consensus within a group,
  • Respect for and understanding of the validity of other viewpoints, and
  • Putting the good of the group ahead of the individual.

Putting together the what, when and who has created the most successful business intelligence programs in the world.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access